Follow-up studies are of particular interest. One of the most important was carried out by Goldfarb who chose two groups of children of similar heredity. Those in the first group had been brought up in institutions until the age of 3 and then placed in the care of foster parents, whereas those in the second group had been handed over to foster parents from the outset. In all cases separation had taken place within the first 9 months of life. The lack of intellectual ability, and particularly the ability to conceptualize, were particularly marked in the group sent to an institution at an early age (22).
On the effects of the length of the intersib interval Anastasi (11) reports on a French study in which
. . . there were 1,244 two-sibling families . . . both siblings had been tested. These were separated into "long interval" and "short interval" sibships, the latter being defined as those falling at or below the median interval. On the intelligence test, the children with long intersib intervals obtained significantly higher means, these differences persisting within each of the five occupational categories into which the sample was subdivided. With long intersib intervals the scores approximated those of only children ówith short intersib intervals, they approximated the scores obtained by 3-child sibships.
It should be stressed that the economic consequences of different levels of nutrition, especially with respect to calories, is probably on a sounder basis than many of the other aspects we have considered. There has been a considerable amount of work on the relationship between the calorie intake and work capacity, and to some degree on the relation between calories and actual output. Unfortunately, the studies involved have not been carried out in less developed countries; therefore, some transference of results is necessary from wartime conditions in advanced countries to the less developed countries. The validity of the transference is to some degree an open question. A good deal of the work is summarized by Keller and Kraut of the Max Planck Institute of Physiology (23). It is of some interest, perhaps, that the relation between calorie change and output change per worker in Germany differed, as we might expect, for different types of work, but the degree of the change for relatively heavy work is quite striking. For example, in a group of coalminers an increase in calories by 33 percent appeared to be associated with a 40 percent increase in output. For steelworkers a 33 percent reduction in calories from an 1,800-calorie level was associated with a slightly larger percentage reduction in output. Although such numbers are at best only suggestive, they nevertheless indicate that at the lower calorie levels, say beginning with 1,800 calories per day and moving downwards, calorie reduc-